Objective For the adaptation of epidemiological characteristics of malaria during its elimination in China and the demand of assessing malaria risk of various areas, a framework for assessing county-level malaria risk needs establishing to provide a basis for malaria surveillance and health resource allocation. Methods The indexes were collected by literature review in the field of malaria risk and on the basis of former studies, a preliminary framework of assessing malaria was established through the panel discussion of the research team. The Delphi method was applied to refine the framework, with the entropy weight method along, and determine the weights of the indexes. Results The experts' active coefficient is 100%, the authority coefficient for each expert is larger than 0.7, and Kendall coefficient is 0.225, χ² = 137.868,P <0.05. The two-tier framework consists of 6 primary indexes and 29 secondary indexes. Primary indexes are risk of high-risk group getting infected, risk of importation, risk of vector receptivity, risk of delay in detecting cases, risk of inefficiency in management of case, and risk of insufficiency in maintaining and guaranteeing capacity to eliminate malaria. Weights for each are 0.169, 0.172, 0.156, 0.171, 0.165 and 0.168 respectively. The five secondary indexes with the highest weights are availability of malaria testing, completeness of the antimalaria drug reserve, the number of imported plasmodium subspecies, intensity of imported malaria, and cross-border migrating population. Weights for each are 0.074 7, 0.069 9, 0.062 7, 0.057 8 and 0.049 9 respectively. Conclusion Results are qualified in responsiveness, authoritativeness, and coordination of experts. Expert consensus and differences among the observations of the indexes are taken into account when determining the weights of the indexes. The framework adapts to the demand of malaria elimination.